A set of Senate races could have a key impact on artificial intelligence regulation and the expansion of data centers.

On Tuesday night, voters in Ohio and Texas tapped former Sen. Sherrod Brown and fellow Democratic candidate James Talarico for Senate spots, according to the Associated Press.

In Ohio, incumbent Sen. Jon Husted was viewed as taking a friendlier stance on data centers compared to former Sen. Brown, who has been more staunchly opposed to them.

In Texas, candidate Talarico has said he would rein in data centers and would ensure that AI centers don't raise energy costs. Meanwhile, Texas Gov. Greg Abbott previously encouraged AI companies to develop in the Lone Star state, but has since reportedly shifted his stance, placing a temporary moratorium on data centers until the end of the year.

In Michigan, U.S. Senate candidate Abdul El-Sayed has said he would push for "public ownership of AI," while his Republican challenger Mike Rogers has taken a more business-friendly approach to data centers. That race, as of late Tuesday night, had not been called.

Artificial intelligence policy has become a major flashpoint amid concerns around safety and environmental risks. For the latter, the build-out of data centers has sparked anxieties between both Republicans and Democrats alike over rising costs for energy bills to residents and environmental impacts. The Federal Reserve has also been examining how AI's productivity could have an outsized effect on the economy given that it is stoking short-term demand and putting price pressures on electricity and construction.

Half of people in the U.S. oppose having data centers built within three miles of their home and are more likely to go against candidates who are pro data center construction, according to results published on Oct. 25 by The POLITICO Poll, conducted by PF Global.

As for safety, hacks over the summer, including one where OpenAI models escaped their controlled environment and hacked a company called Hugging Face, raised concerns over how effectively AI models can be controlled. In September, Jacob Coxon, a researcher who had worked at both OpenAI and Anthropic, resigned from Anthropic and publicly accused the companies of moving too quickly toward AI, saying that they were "racing straight to self-improving superintelligence and gambling with our lives."